X (Twitter) Buying Signals for B2B: 9 Searches to Run

X (Twitter) buying signals for B2B are public posts where someone asks for a tool, complains about the one they have, compares two vendors, or announces a change (funding, hiring, a launch) that creates a need. You find them with X's advanced search, which works only when you are logged in, using the nine searches below. Then you reply in public, match the handle to a LinkedIn profile, and only then send a message. Fifteen minutes a day covers most categories.

What counts as a buying signal on X

A buying signal is a public action that shows a person is closer to buying than they were yesterday. On X the strongest ones are written in the first person and ask for something: "anyone recommend a tool for", "we're switching from", "is there an alternative to". Weaker ones are context: a founder announcing a round, a company hiring for the role your product serves, someone replying to a competitor's support account.

The test is the same one used in the lesson on what a buying signal is: does the post tell you the person has the problem you solve, and does it tell you they have it now? A post that passes both is worth a reply today. A post that passes one is worth a note in a list. A post that passes neither (a like, a follow, a retweet of an industry article) is activity, and chasing it is where most X prospecting time goes to waste.

X differs from LinkedIn in two ways that matter for this work. People write more candidly on X, so complaints and comparisons are blunter and easier to find. But an X profile rarely states company and role, so a signal on X gives you a problem and a handle, and you still have to work out who is behind it. That second step is covered below, and it is the one the pages ranking for this query skip.

Nine X (Twitter) searches that find B2B buying signals

Type each search into the box at the top of x.com, or build it in the advanced search form, then open the Latest tab so results sort by time. Replace the words in capitals. X's help page covers exact phrases, any-of and excluded words, hashtags, language, posts from, to or mentioning an account, and date ranges; the form writes those operators for you. The engagement and content filters (-filter:replies, filter:links, min_faves:, ?) are not on X's help page; they are listed in Typefully's operator guide, and X can change them without notice, so check that each returns results before you rely on it.

1. The recommendation ask. ("anyone recommend" OR "any recommendations" OR "what do you use for") CATEGORY -filter:links -filter:replies The strongest signal on the platform. The link filter removes vendors posting their own listicles.

2. The alternative search. ("alternative to COMPETITOR" OR "COMPETITOR alternative" OR "switching from COMPETITOR" OR "moving off COMPETITOR") -from:COMPETITOR Excluding the competitor's own account removes their marketing and leaves their customers.

3. The complaint. COMPETITOR ("so slow" OR "keeps breaking" OR "price increase" OR "doubled our" OR "cancelled" OR "cancelling") -from:COMPETITOR -filter:links Complaints are candid on X. Read the thread before replying; some are resolved in the replies.

4. Replies to the competitor's support account. to:COMPETITORSUPPORT -from:COMPETITORSUPPORT Everything people write at a vendor's support handle is a problem they have today. Add ? to keep only questions.

5. The comparison. (COMPETITOR_A vs COMPETITOR_B) OR ("COMPETITOR_A or COMPETITOR_B") -filter:links The person is choosing this week. If you are a third option, say in one line why, and when you are not.

6. The milestone. ("we just raised" OR "we raised" OR "closed our seed" OR "closed our series") -filter:replies since:2026-10-29 Funding is context, not intent. It becomes a signal when your product is what a company buys after that round. The since: date keeps it to the last week; move it forward each Monday.

7. The hiring post. ("we're hiring" OR "we are hiring" OR "looking to hire") ROLE -filter:links A company hiring its first SDR, first RevOps lead or first designer is about to buy the tools that role uses. The hiring signals routine shows how to run this weekly across job boards and LinkedIn as well.

8. The question with reach. CATEGORY ? min_faves:20 -filter:replies Questions that already drew likes have an audience, so a good public reply gets read by more than the asker. Lower the number for narrow categories.

9. Inside a List. list:YOURHANDLE/LISTNAME ("looking for" OR "recommend" OR "alternative") Once you have a List of accounts that fit your ideal customer profile (next section), this searches only their posts, which removes almost all the noise of the open searches above.

Save each search as a bookmark in your browser with the Latest tab selected. X can save a search but does not tell you when it has new results, so the bookmark folder, opened at the same time each day, is the alert.

How do you use X Lists to cut the noise?

Open searches on X return students, vendors, journalists and bots alongside buyers. A List fixes that. According to X's own guide to Lists, a List defaults to public, and turning on the Private slider makes it "only accessible to you". Names are limited to 25 characters and cannot start with a number. Keep lead Lists private; nobody needs to know they are on your prospect list.

Build three. First, competitor customers: people who have mentioned a competitor by name in the last six months (search 2 and 3 above, then add each author). Second, in-profile accounts: founders and operators whose bios match your ideal customer profile, found by searching the job titles your buyers use with filter:links removed and the bio read by hand. Third, loud accounts in your category: the ten or twenty people whose posts your buyers reply to, because the reply threads under those posts are where recommendation asks and complaints pile up.

Then search 9 runs across all three. Fifteen minutes on the List timelines each morning, newest first, replaces an hour of open searching.

What do you say, and where?

Reply in public first. A useful reply under a recommendation ask is read by the asker and by everyone who finds the thread later, and it costs you nothing in permissions. Answer the question as asked; if your product is the honest answer, name it in one clause and say who it is not for. If someone else's product is the honest answer, say that, and you will be remembered the next time.

Direct messages come second, and X decides whether you can send one. According to X's Direct Message FAQ, a person you do not follow can message you first only if your "Receive Direct Messages from anyone" setting is on; otherwise you have to message or follow them before they can reply. The same gate applies in reverse: if the prospect has that setting off and does not follow you, the DM button will not be there. Check before you write the message, and when it is missing, the public reply is your only channel on X.

Keep the DM to three sentences. First, the post you are replying to, quoted or paraphrased, so they know this is not a template. Second, what you did about the same problem, one concrete line. Third, a question that is easy to answer, not a meeting request. The LinkedIn templates for buying signals carry over to X with the formality turned down; X readers expect shorter, plainer messages.

Volume is the thing to get right, because X's rules are explicit about it. X's platform manipulation and spam policy prohibits "sending bulk, aggressive, high-volume unsolicited replies, mentions, or Direct Messages", "repeatedly sending identical Direct Messages", and "aggressively or automatically engaging with Posts to drive traffic or attention to accounts, websites, products, services, or initiatives". A handful of hand-written replies and messages a day to people who asked a question in public is ordinary use of the site. A script that replies to every match of search 1 is the thing the policy describes.

Why the handle is not yet a lead

Most X profiles give a first name, a bio written for an audience, and a link. They do not reliably give the company, the role, the company size or the country, which are the four things you need before anyone spends time on the lead. A recommendation ask from a founder of a twelve-person SaaS company and the same words from a student doing a project look identical in search results.

So the step after the public reply is identification. Open the profile and follow the link in the bio; most operators link their company site or their LinkedIn. If the link is a personal site, the about page usually names the company. Then search the name and company on LinkedIn and read the profile: title, headcount, location, whether they match the profile you sell to. The qualification checklist in how to qualify leads before outreach is the same one to run here, and if the person does not pass it, the public reply was still worth leaving.

When they do pass, LinkedIn is usually the better channel for the follow-up, because the person's role is visible there and the conversation does not depend on an open-DM setting. This is also where the manual part of the work ends for us: Intentsly watches the same kind of signal on LinkedIn (people posting about the problem you solve, or engaging with competitor and industry-leader posts), scores each person against your ideal customer profile and can run the conversation through its AI SDR on LinkedIn. It does not monitor X, which is why this article stops at the LinkedIn handoff rather than promising an X feed.

Can you automate the monitoring?

The reading, yes. The replying, no, for the policy reasons above. There are two ways to automate the reading, and both cost money, which is worth knowing before you choose.

The first is X's own API. X's developer pricing page describes a pay-per-usage model with post reads at $0.005 per resource, user reads at $0.010, and a cap of 3 million post reads per monthly billing cycle, above which it points to Enterprise. A small script that runs searches 1 to 5 every hour reads a few thousand posts a day, so the bill is modest, but someone has to write and maintain the script and the keyword list.

The second is a keyword-alert tool. Buska's X page, opened on 5 November 2026, says X is included on every plan, at $49 a month for 5 signals scanned daily, $99 a month for 15 signals every 3 hours, and $249 a month for 30 signals hourly, with a 7-day trial and no card; it says it searches X directly rather than through the API, drops posts older than 90 days, and scores each match from 0 to 100 for intent. Prices change, so the link is the source, not this paragraph. The Reddit equivalents, with their own prices and a free method, are compared in Reddit keyword monitoring for leads.

Whichever you pick, start with the bookmark folder for two weeks. If the nine searches produce fewer than a handful of real signals a week in your category, no tool will change that, and you have saved the subscription.

When X is not worth the time

Run searches 1, 2 and 3 for your category and your two main competitors, sorted by Latest, and count the posts from the last 30 days written by people who could plausibly buy. Then apply the rule:

• Under five in a month. Your buyers are not asking on X. Leave the bookmarks, check them monthly, and spend the daily fifteen minutes on LinkedIn, where the LinkedIn buying signals for the same categories are usually denser and the person is identifiable from the start.
• Five to twenty a month. X earns fifteen minutes a day, by hand, with the three Lists. Categories whose buyers build or market in public (developer tools, SaaS, marketing) tend to land here, but count before you trust that.
• More than twenty a month. Automate the reading with a tool or the API, keep the replying by hand, and give one person the job of working the feed first thing each morning, because the first useful reply under a recommendation ask is the one that gets the DM.

The categories where X rarely clears the first bar are the ones whose buyers are regulated, offline or senior enough not to post in public: manufacturing, healthcare administration, legal, most of enterprise IT. For those, the same signals exist on LinkedIn and in hiring data, and X is a distraction.

What to do next

Run search 1 for your category right now; it takes a minute and tells you whether X is a channel for you at all. If it is, build the first private List today and bookmark the nine searches with Latest selected. For the LinkedIn side of the same buyers, try Intentsly's signal feed to see who is posting about your problem this week with no account needed; when we launched, our own 14 qualified demos in 14 days came from LinkedIn signals caught in real time, with zero manual outreach, and that is the part of this routine you do not have to do by hand.

FAQ

What counts as a buying signal on X?

A public post that shows the person has the problem you solve and has it now: asking for a recommendation, naming an alternative they want, complaining about a vendor, comparing two tools, or writing at a competitor's support account. Funding and hiring posts are context that becomes a signal when your product is what that company buys next. Likes and follows are activity, not signals.

Can I search X without an account?

Not with advanced search. X's help page says advanced search is available when you are logged in to X.com, and the operator-based searches above also need a logged-in session. That page does not name Premium as a requirement. The limits you will hit are noise and time, which the private Lists are there to reduce.

Should I reply publicly or send a DM?

Public first, always. The reply is read by the asker and by everyone who finds the thread, and it needs no permission. DM only after a useful reply, only if the DM button is there (it depends on the other person's settings), and only in three short sentences that quote their post. Never send the same DM twice; X's spam policy names repeated identical messages as a violation.

Can I automate X buying-signal monitoring?

You can automate the reading, through X's pay-per-usage API or a keyword-alert tool, both priced on their own pages. You cannot automate the replying or the messaging without breaking X's platform manipulation policy, which prohibits bulk or automated replies, mentions and DMs. Keep every reply written by a person, and keep the daily number small.

Does X still work for B2B lead generation?

For some categories. Where founders and operators post in public (developer tools, SaaS, marketing), the recommendation and complaint searches tend to return real buyers. Where buyers are regulated or senior enough not to post, the searches come back empty and LinkedIn is the better use of the time. The count rule above tells you which you are in within an hour.

How do I turn an X handle into a qualified lead?

Open the profile, follow the bio link to the company or personal site, then find the person on LinkedIn and read title, headcount and location against your ideal customer profile. If they pass, continue the conversation on LinkedIn, where the role is visible and messaging does not depend on an open-DM setting. If they do not pass, your public reply was still a useful answer, and it cost a minute.